Independent Wheelchair Transfers in the Built Environment: How Transfer Setup Impacts Performance Phase 2: Final Report

Implication for Standards and Design

Close to 200 community dwelling adults who independently transfer have been evaluated during phases 1 and 2 combined of this multi-year study.  The main difference between phase 1 and 2 relates to the tool used to collect the data on transfers.  Each phase involved a highly customized transfer station iteratively developed with scientific, expert and consumer input to support the goals of the project.  The station used in this second phase permitted greater freedom of wheelchair positioning around the transfer surface compared to the first phase station.  Moreover, the second phase transfer station allowed for evaluating the presence/absence of various height grab bars, a backrest added in parallel to the front edge of the transfer surface, a variable width transfer surface, and adjustable height transfer steps.

Transfer Heights

Despite differences in the data collection tools, some results from the second phase of the study were consistent with first phase of the independent transfer study.  For instance, level seat heights were 21.7± 1.2 inches and 22.1±1.4 inches for the first and second phases respectively. Both of these heights are outside of the standards for heights for most elements where adults would be expected to transfer (between 17 and 19 inches [7]).  Also, like the first phase, the second phase found that the majority of WMDs (92% of subjects in the first phase and the 5th percentile (or approximately 94% in this study) could transfer within one inch of the mean wheelchair seat to floor height (22 inches).   Transfers to level surfaces require less exertion on the upper extremities [5, 10, 11]. Transfers are easiest and safest to obtain when they are as close to level as possible [6]. This is consistent with other research results that have found that level transfers require less exertion on the upper limbs [5]. Based on the results of this study, it is estimated that the current standards would accommodate close to the 25th percentile of WMD users (approximately 75%) who can independently transfer to a transfer surface or platform.

The current station tested the ability of the study participants to transfer to an adjustable table height ranging from 10 inches to 43 inches. Although the study did not aim to determine what range of adjustable platform height would be needed to accommodate the majority of WMD users, the data from the subjects in this study (Table 9) suggests a range between 19 to 27.5 inches above the ground or floor surface. This is based on the lowest wheelchair seat to floor height and highest wheelchair seat to seat or floor height of the individuals in the study. All participants can make a level transfer, and the majority of participants can transfer 1 inch above or below their level seat height. Thus adjustability, including this range, would include all of the study participants and allow for greater accessibility where transfers are needed.

Two Step Transfer Evaluation

The evaluation of two-step transfers has never been studied before and was evaluated in this phase. The two-step transfers consisted of an adjacent transfer (no gap between two steps) and a ramped transfer over a constant 8 inch gap between two steps. It should be noted that for this evaluation not all study participants are included in the two-step analysis. Not all participants attempted the two-step transfers as they were given the choice to not make the transfer. Additionally, the study participants who made the two-step transfers appear more able on these transfers when comparing the heights achieved during the two-step transfers to the heights achieved from the WMD to and from the first platform (Table 9).  Transferring to and from a WMD is different than transferring between two surfaces.  It's also possible that more able or skilled WMD users completed the two step transfers.

In this study, participants who attempted the two-step transfers transferred from one step to a second step that was higher and lower than first one. The results of our study suggest that transfer steps of any height differential that currently exist in the built environment would exclude about 15.2% of adult community dwelling WMD users who independently transfer (11 out of 72 participants; 8 of which wouldn’t attempt it at all and 3 of which attempted it but failed in the process).  The addition of a ramp to bridge the two steps did not improve the situation and there were additional failed transfer attempts when the ramp was used to go up a step compared to when the ramp wasn’t used.  It’s possible that more failed attempts occurred because it was ‘new’ and a very different type of transfer than what these individuals were used to.  These types of transfers are not currently part of the standard training that patients receive during a typical rehabilitation stay.  Performance may improve with training and practice. 

Results from the questionnaire asking for participants’ opinion on these transfers reflect a mixed review of the ease and practicality of two-step transfers. Comments on the two-step transfers varied from thinking they were fun, easy, and safe to impractical, difficult, and uncomfortable. Both types of transfers received relatively equal responses for the number of people who said they would or would not perform the transfer in a ‘real-world’ situation. Participants who attempted and were successful with the two-step transfers performed at a high level on average.  The 50th percentile for the adjacent and ramp transfer attained a height of 7.0 inches, only 1 inch below the maximum allowable transition for this transfer (8.0 inches). A lower step height (about 2.0 in) would be needed to accommodate the 5th percentile of users who would attempt this type of transfer.

Clear Space Evaluation

nlike [sic] the first phase of the transfer study, subjects in this study were not impeded by any obstacles on the floor that limited the manner of which they approached the transfer station (e.g. subjects in the previous study could not transfer to the station face on).  As a result subjects positioned themselves in a wide variety of locations about the transfer station.  Both the physical size of their WMD, where they positioned it with respect to the station and at what angle they positioned their WMD were all factors that weighed into the space analysis.  Graphical techniques were used to better understand the relationships between these factors.  As can be seen in Figures 7 and 8 more participants set themselves up on the right side of the transfer station compared to directly in front of or to the left of the station.  This may be explained in part by the effects of hand dominance on transfer ability and preferences.  The majority of study participants were right handed and likely their right side was stronger than their left side.  Positioning themselves to the right of the station allows them to use their right side as the trailing arm which carries more of the force during a transfer than does the leading arm when moving toward a new surface.  

Additionally, from Figure 7 it can be seen that some study participants did not transfer entirely within the boundaries of the transfer station. The space needed by the study participants was compared to a turning space dimension as described in the ADA-ABA standards (60x60 inch space). All of the study participants were able to transfer within the 60 inch depth, in fact the largest depth dimension required by anyone subject was 55.5 inches.  However, when looking at Protocol A for the maximum height transfer only 65% of the population could transfer entirely inside of the 60 x 55.5 inch defined space. For the transfer to a lower platform that number decreased to 53% when moving to the station and 56% when coming back to the station. (Protocol A).  A similar trend for participants requiring more space to transfer lower versus higher was found for the other protocols as well (B and C).  The reason that subjects may have used more space to transfer lower than higher may be that they needed more clearance space to accommodate the pivoting swing of the buttocks and increased knee and hip flexion that occurs when moving their body closer towards the ground than when moving their body further from the ground (e.g decreased knee and hip flexion).  Increasing the width of the clear floor space an additional 12 inches to 72 inches would accommodate 75-77% of users who would be expected to transfer higher or lower than their seat to floor height.  In order to accommodate 95% of the study population transferring higher and lower than their mobility device seat, the total width would need to be increased to 92 inches total.

As mentioned previously, the right side of the transfer station was preferred by the study participants. Shifting the clear floor square area requirement more towards the right may accommodate more people and minimize the overall amount of space required around transfer elements.   Like the previous study, subjects were not restricted to transferring from a certain direction (e.g. to their left or right) or within certain predefined clear floor spaces.  This was done to better assess limitations associated with transfer heights and to limit the number of transfers that each subject needed to perform.  Had subjects been limited by a certain direction or standard space dimension, it’s possible that fewer subjects could have completed the transfers at all or as high or as low as they did due to a lack of function or inadequate space available to position their WMD and themselves. 

For the angle of WMD positioning, participants had the option to align themselves parallel to the station (at 0 degrees), perpendicular to the station (at 90 degrees) or at any angle in between. When transferring to and from the station, the 50th percentile study participants preferred an orientation angle of around 30 degrees. It’s important to note that the angle of positioning can increase the amount of space required (e.g. in either the width or length dimensions) and should be accounted for in the future when planning the amount of clear floor space to design for around transfer elements. 

Seat Width Evaluation

An 18 inch seat was the smallest of the platform seat widths available in this study and was suitable for many participants (the 25th through 95th percentiles) for their level, highest and lowest height transfer regardless of the presence or absence of grab bars or a backrest on the transfer surfaces.  The 5th percentile wheelchair users however required transfer platform width of 25 inches if there were grab bars on the platform and 27 inches without the grabs bars.  The additional two inches for the simple platform transfer was likely needed to allow hand placement or grip directly on the transfer surface. The additional platform width may also have been needed to accommodate larger individuals; however the size of the individual and seat width of the wheelchair were not measured and therefore cannot be used to explain the transfer platform widths selected by the participants in the study. Seat width preference did not change much over the protocols. It was felt that some participants would have had greater performance on the station if they had opted for a larger seat width and more room to position the hands for transfer, but since they did not ask to increase the width when the option was offered it was left the same. It may be of value in future work to investigate transfer performance when the seat width is forced to vary between protocols and transfer trials. This was not done in the current study to minimize the number of transfers performed.

Grab Bar and Backrest Evaluation

Grab bar and backrest preferences were reported in this study with the tall grab bar and the short backrest being selected more often than the short grab bars and medium and tall backrests. The grab bars were frequently used as handhelds during the transfer process and for one person it made the difference of being able to do a non-level transfer to the platform. The analysis also revealed that the presence of grab bars helped participants to transfer higher and lower than their mobility device seats. Grab bars should be placed everywhere persons would encounter a non-level transfer. This would allow persons with disabilities to make non-level transfers more easily and safely. The backrest was used as a handheld less frequently and more often when participants were transferring lower versus higher. The backrest may have helped to provide a sense of security, comfort or support for participants when seated on the platform but it did not appear to be as useful as the grab bars were for assisting with the transfer. One possible explanation for this is that unlike the phase 1 study, participants were able to transfer from the front of the station as well as from the side of the station. In the situation where a participant transferred from the front, they may not have needed the backrest or the backrest may not have been within reach.

Secondary data analyses were performed to compare phase 1 and phase 2 results. A comparison of participant demographics and transfer performances between the two phases can be found in Addendum D [Note that Addendum D was a preliminary analysis that was done before completion of the phase 2 study and thus contain a subset of the total numbers of participants tested]. Also, a comparative analysis was done to investigate differences between the veteran population that was tested at the National Disabled Veterans Winter Sports Clinic and the population that was tested at the Human Engineering Research Laboratory. This analysis was performed using data from phase 2 and the results are included in Addendum E [Note that Addendum E was a preliminary analysis that was done before completion of the phase 2 study and thus contain a subset of the total numbers of participants tested]. Lastly, a comparative analysis looking at the difference in transfer abilities between men and women can be seen in Addendum G.

Sample Size Estimation

A sample size estimation was performed to estimate the number of participants that would be needed in a future study. Data from the maximum and minimum height transfers from protocol A were used for this analysis. The percentile minimum and maximum heights attained for all 71 subjects are shown in Table 9.  Ten participants were selected at random from the 71 participants who performed transfers to a higher and a lower height seat. We then added 10 more participants from the existing data set (randomly selected from the total sample) and reassessed the values again.  This process was repeated until the data for all 71 participants were included in the analysis (last rows, Tables 20 and 21).

Table 20. Percentiles for Lowest Attainable Transfer heights Protocol A

Table 21. Percentiles for Highest Attainable Transfer heights Protocol A

The estimates for the highest and lowest attainable heights across all percentiles became relatively stable with 60 participants. This analysis suggests that our sample may have reached a potential saturation point at or around these numbers and that adding additional adult participants with similar characteristics to the adults in this study would not change these results. However, adding additional participants to the study with different demographics than the ones tested may adjust these values.

Study Limitations

Some limitations that occurred with this study were that occasionally the grab bars would get in the way of a participant transferring to or from the station since they were located on either side of the transfer surface. These participants were observed having to reposition their WMD or having to transfer around the grab bar. A protocol or an option of using only one grab bar or having a removable grab bar may be worth investigating in the future. Also, the grab bars were placed perpendicular to the front edge of the transfer station. This may have forced some participants to approach the station more from the front (e.g. forward-facing) than from the side. However, the study participants on average did not appear to adjust their angle of approach too much in between protocols with and without the grab bars present on the station (Table 14).     
 
Some participants also asked if they could use both size of grab bars on either platform during the two-step transfers. They mentioned that the taller ones were more helpful for going to lower elevations and the lower ones for transferring higher. This was not an option for this study and it may be that developing a grab bar that is angled or easily adjustable in height would be beneficial for future work. There was not a backrest available behind the second platform for the two-step transfers. When performing these transfers to a higher elevation participants could be transferring to a seat around 31.5 inches off the ground. Even though there were spotters standing behind this platform having a wall, backrest, or another step available there would have added to the safety of the transfer. The grab bars attached to this surface did help with keeping the participants stable by allowing another surface to grip onto.

In addition to the grab bars, the backrest can also be seen as a study limitation. In this phase of research, the backrest was placed parallel to the front edge of the first transfer platform whereas in phase 1 the backrest was placed perpendicular to the front edge of the transfer platform.  This phase 2 backrest placement meant that participants were not able to orient themselves right next to the back rest but instead would need to reach across the platform if wanting to use the backrest as a handheld. For many participants, the backrest was too far out of reach for them to grab it which would explain why very few participants placed their hands on the backrest to help them transfer. 

Another limitation concerns the small sample size collected. Although the above sample size analysis suggests that we may have reached a saturation point, for standards research this is still considered a small sample size and may not reflect the demographics of the cohort of individuals who independently transfer. Additional study participants would add to the strength of the study and help to insure the study sample is as representative of the cohort population as possible.

The study was open to children seven years old and older to participate. The Access Board was particularly interested in collecting data from school aged WMD users to provide guidance on playground equipment standards. Extensive efforts were made to advertise the study to different schools, hospitals, and organizations known to work with children and only one family over the course of the study had contacted us to participate.  This family's son was eligible for the study but his parents were not able to bring the child to HERL to be evaluated.  We were able to collect data from young and small adults. The youngest participant was 18 years old and four participants were less than 4.9 feet tall.

Another study limitation was the low ratio of women compared to men in the study (23% of the participants were women). Every effort was made to recruit women into the study and the proportion of women examined in this study is similar to other studies that have involved community dwelling adult wheelchair users [6, 12-14].  There is a lack of data on the numbers of women and men who use WMDs and who independently transfer in the community. Another issue concerning the US census statistic on the numbers of women versus men WMDs is that it does not take into consideration the disparities associated with age and mobility use.  Kaye et al. states that there is a dramatic increase in WMD use with age [15]. Only 0.41% of the general population uses a WMD from the ages of 18 to 64 years, but that percentage increases to 2.99% for the population 65 years or older. It's possible that due to a women's longer life expectancy that the census statistic is skewed towards including more older women (over 65) who would not likely be actively transferring out in the community (e.g. who are residing in institutional settings).  Moreover data on working age adults shows that more men use mobility devices in general than women (1.6% of men vs 1.3% of women of the population) [15].  Our study involved men and women who were mostly working age and thus it’s possible that the true proportion of women to men who actively transfer in the community is closer to that represented in this study.